MOLECULAR CYTOGENETICS OF RENAL CELL TUMORS Gyula Kovacs* National Cancer Center Research Institute, Tokyo, Japan
I. Introduction 11. Differential Genetics of Renal Cell Tumors 111. Genetics of Nonpapillary Renal Cell Carcinomas A. Sporadic Renal Cell Carcinoma B. Familial Renal Cell Carcinoma C. Renal Cell Carcinoma Associated with von Hippel-Lindau Disease D. Nonhomologous Mitotic Recombination E. Somatic Mosaicism in Normal Kidney Tissue F. Model of Development and Progression of Nonpapillary Renal Cell Carcinoma IV. Genetics of Papillary Renal Cell Tumors A. Genetic Alterations Associated with Papillary Renal Cell Tumors B. Adenoma-Carcinoma Sequence C. Parenchymal Lesions Associated with Papillary Renal Cell Tumors D. Shared Genetic Changes in Wilms’ Tumor and Papillary Renal Cell Tumor E. From Developmental Disturbances to Papillary Renal Cell Carcinomas F. Papillary Renal Cell Tumors with Translocation Involving the X p l l . 2 Breakpoint V. Renal Oncocytoma A. Chromosomal Alterations B. Mitochondria1 DNA Alterations VI. Chromophobe Renal Cell Carcinoma A. Chromosomal Aberrations B. Mitochondria1 DNA Alterations VII. Conclusions References
I. Introduction T h e development and progression of cancer is associated with alterations of genes that control growth and differentiation. The accumulation of genetic changes, such as inactivation of tumor suppressor genes, and the alteration of function of oncogenes, growth factors, and their receptors, as well as genes related to metastatic growth characterize the *Present Address: Institute of Neuropathology, University Hospital, CH-809 1 Zurich, Switzerland. 89 ADVANCES I N CANCER RESEARCH. VOL 62
Copyright 0 1993 by Academic Press, Inc. All rights of reproduction in m y form reserved.
90
GYULA KOVACS
multistep nature of tumor development (Fearon and Vogelstein, 1990). The mutational inactivation of one allele and the loss of the wild-type allele of tumor suppressor gene is a well-characterized sequence of genetic events associated with the development of cancer (Knudson, 1987). Recent genetic studies are focused on detection of loss of chromosomal segments that may harbor loci of tumor suppressor genes. Allelotyping, i.e., using DNA polymorphism to determine the allelic status at each chromosome arm, is now a widely used method to establish specific genetic changes in cancer (Vogelstein et al., 1989). However, solid tumors are characterized not only by loss of chromosomal segments but also by trisomies or partial trisomies and by translocations. A balanced translocation could not be detected by allelotyping, and a trisomy might be easily overlooked and designed as loss of heterozygosity by restriction fragment length polymorphism (RFLP) analysis. Using cytogenetic methods, one can recognize loss of chromosomal segments, trisomies, and translocations. T h e aims of this review are threefold: First, I show that cytogenetic analysis, especially in combination with DNA analysis, remains a powerful technique to detect new tumor-associated genetic alterations; second, I show that molecular cytogenic methods are efficient to stratify renal cell tumors; third, I will try to emphasize the complexity of genetic alterations in subtypes of renal cell tumors and indicate that specific genetic changes are useful in the diagnosis as well as in the separation of high-risk groups for therapy. II. Differential Genetics of Renal Cell Tumors Renal cell carcinoma is the most common malignant tumor arising from the kidney and affects about 7 of 100,000 adults (Javadpour, 1984). Epidemiological studies indicate that renal cell carcinoma is somewhat more common in Scandinavians and North Americans and that the incidence is lower in Asians and Africans. Although a moderate association between tobacco use and the incidence of renal cell carcinomas has been described, no conclusive evidence could be established for the role of environmental effects in their development (Bennington and Labscher, 1968). T h e vast majority of renal cell tumors occur in sporadic form. Predisposition to tumor development was described only in rare families and individuals with von Hippel-Lindau disease (Cohen et al., 1979; Kovacs etal., 1989a; Lamiell et al., 1989; Li et al., 1982). There is no satisfactory method for an early detection of renal cell tumors, and almost 40% of the patients have a metastatic tumor at the time of diagnosis. T h e most effective therapy for renal cell carcinoma localized to
MOLECULAR CYTOGENETICS OF RENAL CELL TUMORS
91
the kidney is the surgery, whereas a metastatic tumor is practically incurable. The overall response to biological response modifiers is low, and the treatment is only palliative. The goal of future therapy is to target tumor suppressor genes. Therefore, it is important that genetics be incorporated into the evaluation of renal cell neoplasm in an attempt to provide a foundation for future diagnosis and selective treatment. In recent years, a large number of cytogenetic and RFLP studies were carried out to detect specific chromosomal aberrations in renal cell carcinomas. Although the molecular basis of genetic changes is not yet established, the combination of such chromosomal and DNA alterations stratify distinct subtypes of kidney tumors (Table I). We must be constantly aware that we are not dealing with a single disease, called renal cell carcinoma, but with genetically well-characterized types of tumors, each with a unique natural history. The genetic changes, which are associated with tumor progression, may help to estimate the biological behavior of renal cell carcinomas. This classification is very recent and still not widely accepted (Kovacs, 1990). However, it is important to understand that these tumors develop on the basis of separate molecular mechanism and do not have an association with each other. Nonpapillary renal cell carcinomas show a loss of chromosome 3p segments in a proportion of tumors similar to the frequency with which the Philadelphia chromosome is observed in chronic myelogeneous leukemia. Papillary renal cell tumors have a genetic marker in the form of trisomy 17, a genetic change that has not been found in nonpapillary renal cell carcinomas. None of the papillary renal cell tumors shows a rearrangement of chromosome 3p or 5q segments (Kovacs et al., 1989~). Renal oncocytomas and chromophobe renal cell carcinomas have characteristic chromosomal and mitochondrial DNA alterations.
Ill. Genetics of Nonpapillary Renal Cell Carcinomas A. SPORADIC RENAL CELLCARCINOMA Sporadic nonpapillary renal cell carcinomas account for about 80% of the renal cell tumors (Kovacs, 1993). Histologically, they display solid, trabecular, tubular, o r cystic growth pattern. Although most of these tumors are made up of clear cells, large areas or the entire tumor may be composed of granular cells. The incidence of nonpapillary RCCs is about 1.5-2 times higher in males than in females. Until now, more than 400 renal cell tumors have been processed for chromosome analysis. After a critical reevaluation of the published data, fewer than 200 cases remain for the present review.
TABLE 1 DIFFERENTIAL GENETICS OF RENALCELLTUMORS Tumor type pRCA" pRCC npRCC chRCC
RO
Alterations of chromosomal (%) and mitochondria1 D N A
-Y 77 93 26 -
+7
+ I 7 +3q
100 100 75 80 18 -
-
34 -
t 8 + I 2 t 1 6 t 2 0 -3p
- -
-
-
18 -
62 -
28 -
- -
70
14
-8p
22 25
-9
-
-14q
-
-1
- - 100 Normal/abnormal karyotypes; translocation ( 1 lq13;?); -Y,34 10
96 25
+5q -6q
14 18
15 41
-2
-6
- -
- - 95 88 1.
-10
-13
-17 ~
-
88
95
76
-21
mtDNA
88
-
~~-
-
+
+
,, pRCA, papillary renal cell adenorna; pRW, papillary renal cell carcinoma: npRCC, nonpapillary renal cell carcinoma; chRCC, chromophobe renal cell carcinoma; RO, renal oncocytoma.
MOLECULAR CYTOGENETICS OF RENAL CELL TUMORS
93
1. Loss of Chromosome 3p Segment The deletion of chromosome 3p segment was detected in the karyotypes of about 30-50% of the renal cell carcinomas in most cytogenetic studies (Berger et al., 1986; DalCin et al., 1988; deJong et al., 1988; Limon et al., 1990; Maloney et al., 1991; Miles et al., 1988; Teyssier and Ferre, 1990; Walter et al., 1989; Yoshida et al., 1986). However, using an appropriate cell culture technique and excluding papillary and chromophobe renal cell tumors and oncocytomas from the series, more than 90% of the nonpapillary renal cell carcinomas display the loss of short arm of chromosome 3 (Carroll et al., 1987; Kovacs et al., 198710; Kovacs and Frisch, 1989; Presti et al., 1991). The chromosome 3p rearrangement is the solely karyotype change in about 10% of these tumors. The RFLP analysis revealed loss of constitutional heterozygosity at chromosome 3p in 53-10076 of renal cell carcinomas (Anglard et al., 1991; Bergenheim et al., 1989; Kovacs et al., 1988a; Ogawa et al., 1992; van der Hout et al., 1991b; Morita et al., 1991; Zbar et al., 1987). These data suggest that alteration of a tumor suppressor gene at chromosome 3p region is associated with the development of nonpapillary renal cell carcinomas. To test this hypothesis, single chromosomes containing 3p, 11, or X chromosomal segments were introduced into the human renal cell carcinoma cell line YCR via microcell fusion (Shimizu et al., 1990). As expected, the complementation with chromosome 3p segment resulted in modulation of tumor growth, whereas introduction of other chromosomes did not change the biological behavior of the YCR cell line. The locus of the putative suppressor gene is not yet determined. An interstitial deletion at 3p13-p21 or at 3p14-p23 segment was described in most cytogenetic studies. Other investigators described an unbalanced translocation between chromosome 3p13 and 5q22, as the most common alteration in nonpapillary renal cell carcinomas (Kovacs et al., 1987b; Kovacs and Frisch, 1989; Kovacs and Kung, 1991; Presti et al., 1991), which might easily be “misread” as an interstitial deletion in chromosome preparations of poor quality. Thus, high-resolution chromosome analysis of tumor cells suggests that the 3p13-pter segment is the smallest overlapping deletion in nonpapillary renal cell carcinomas. RFLP analysis of renal cell carcinomas suggests that at least two loci on chromosome 3p-one at 3p14 and one at 3p21.3-are involved in interstitial deletions (Yamakawa et al., 1991). Other investigators found terminal and interstitial deletions distal to chromosomal band 3p2 1.2 (Anglard et al., 1991; Bergenheim et al., 1989) or an interstitial deletion between chromosomal bands 3 ~ 2 1 . 3and 31324 (van der Hout et al., 1991a). Recently, candidate tumor suppressor genes were isolated from the chro-
94
GYULA KOVACS
mosome 3p21 region. Erlandsson et al. (1990, 1991) identified the acylpeptide hydrolase gene in the vicinity of the D3F15S2 locus, the reduced expression of which was found in renal cell carcinoma tissues. LaForgia et al. (1 99 1) suggested that the protein-tyrosinase phosphatase gamma may be involved in the initiation of renal cell carcinomas and small cell lung carcinomas. T h e lack of expression of aminoacylase-1 gene was shown in small cell lung cancer cell lines (Miller et al., 1989). Although one allele of these genes was deleted, a mutation of the remaining allele was not shown. Either these genes are linked to and deleted with the tumor suppressor gene, or simply they are only one of the genes located to the large chromosome segment deleted in cancer tissues. In most cytogenetic studies, the breakpoint cluster of’ translocation was localized to the chromosome 3p14.2 band, which is the locus of the most common fragile site FRA3B (Sutherland and Hecht, 1985). Fragile sites are located within chromosomal bands, which are frequently involved in rearrangements in cancer cells (Yunis and Soreng, 1984). However, there is no direct evidence that 31314.2 or other fragile sites are involved in the specific chromosomal rearrangements in nonpapillary renal cell carcinomas (Kovacs and Brusa, 1988). The lack of association between fragile site FRA3B and recombinational or deletional breakpoint at chromosome 3p in renal cell carcinomas is also supported by the analysis of induced chromosome fragility of tumor cells carrying a 3p deletion (Tajara et al., 1988). High-resolution chromosome analysis localized the most distal breakpoint to the border of chromosome 3p13 and 3p14.1 chromosomal bands, which is also the site of breakpoint cluster in nonpapillary renal cell carcinomas (Kovacs et al., 1987b, 1988a; Presti et al., 1991). That this chromosomal band represents a “hot spot” in various tumors is suggested by the finding of a submicroscopic homozygous deletion, which involves the D3S3 locus in DNA of the small cell lung cancer cell line U2020 (Rabbits et al., 1990). Although the deletion was localized by genetic and physical mapping to the chromosome 3p12 band (Latif et al., 1992), a deletion mapping of renal cell carcinomas obtained from individuals with constitutional t(3;6) and t(3;8) has placed the D3S3 locus to the 3p13-p14.1 segment spanning the two breakpoints (unpublished observations, 1992). Thus, the homozygous deletion in the U2020 cell line and the breakpoint cluster in nonpapillary renal cell carcinomas are localized to the same chromosomal region. 2. Trisomy of Chromosome 59 Segment
Chromosome 5q22 sequences are frequently rearranged in a specific manner in renal cell carcinomas (Kovacs et al., 1987b), but to date only one paper has paid attention to this genetic change (Presti et al., 1991).
MOLECULAR CYTOGENETICS OF RENAL CELL TUMORS
95
In a large series of karyotype analyses, 20 of 75 nonpapillary RCCs showed an unbalanced t(3;5) leading to monosomy of chromosome 3p and trisomy of chromosome 5q segments, 15 tumors had a trisomy 5, and 3 tumors had a translocation between chromosome 5q22 and chromosomal regions other than 3p (Kovacs and Frisch, 1989). The net result of these alterations is a partial trisomy of chromosome 5q22-qter segment in 50% of the cases. The duplication of one allele of chromosome 5q was confirmed by RFLP analysis of tumor tissues from patients with hereditary and sporadic cancer (Kovacs and Kung, 1991). Recently, the loss of heterozygosity at chromosome 5q2 1 was demonstrated in 33% of renal cell tumors (Morita et aZ., 1991). To establish the nature of chromosome 5q alteration, we employed the RFLP technique for analysis of tumor cells, which were karyotyped for the number of chromosome 5q segments (unpublished observations, 1992). Using the same polymorphic DNA markers that Morita et al. (1991) used, we confirmed the duplication of one homolog of 5q22-qter sequences corresponding to prior cytogenetic findings. In addition, we found the duplication of one allele of loci at chromosome 5q22 in 7 out of 17 carcinomas, which have two normal appearing chromosome 5. We could not confirm loss of heterozygosity at chromosome 5q21-22 in any of the 75 tumors analyzed. Taking into account the results of cytogenetic and molecular studies, the chromosome 5q22 band is affected by allelic duplication in about 70% of nonpapillary renal cell carcinomas. This chromosomal band is the site of a breakpoint cluster in mitotic recombination between chromosome 5q and other chromosomes leading to the partial trisomy of 5q22-qter. There are many growth factors and receptors located next to the trisomic chromosome 5q22-qter segment, the altered dosage and overexpression of which might provide proliferative advantage for tumors cells (Warrington et aZ.,1992). However, the duplication of only a small fragment at the chromosome 5q22 band in many tumors excludes this possibility. The APC and MCC genes are localized to this chromosomal site (Kinzler et al., 1991a,b; Groden et al., 1991), but they are not involved in partial trisomies in all cases. Likely, alteration of a gene or a breakpoint cluster region distal to the APC and MCC genes is the specific event associated with the development of nonpapillary renal cell carcinomas. It might be that a tumor suppressor gene is located at this region, disruption or transpositional inactivation of which is important for tumor development. 3 . Monosomy of Chromosome 6q, Sp, 9, and 14q
A monosomy or deletion of chromosome 14q occurs in about 3050% of nonpapillary renal cell carcinomas (deJong et al., 1988; Kovacs et
96
GYULA KOVACS
al., 1987b; Kovacs and Frisch, 1989; Maloney et al., 1991; Presti et al., 1991; Walter et al., 1989). The chromosome 14q22-qter segment represents the smallest overlapping deletion. Other nonrandom alterations such as loss of chromosome 6q23-qter and 8pll-pter segments and monosomy 9 occurs in 14, 22, and 14% of the cases, respectively. Aiteration of these chromosomal regions are implicated in the genetics of various types of tumors as well.
B. FAMILIAL RENALCELLCARCINOMA Susceptibility to renal cell carcinoma was found in rare families with normal karyotype (Li et al., 1982; Pathak et al., 1982). The association between renal cell carcinoma development and inherited constitutional translocation 3;8 was reported (Cohen et al., 1979). Each member of this family carrying a balanced translocation developed multiple and/or bilateral renal cell carcinomas at an earlier age of onset. It was suggested that one allele of a putative tumor suppressor gene is disrupted by translocation. According to Knudson’s two-hits model, the second hit should affect the normal, nontranslocated chromosome 3p (Knudson, 1987). However, no tumor tissue was available to test this hypothesis. The association between constitutional translocation 3;6 and development of multiple and bilateral renal cell carcinomas was recently reported (Kovacs et al., 1989a). Contrary to the predictions, the normal chromosome 3 homolog was retained and the derivative chromosome 6 carrying the translocated 3p 13-pter segment was lost in tumor tissues. Very recently, cytogenetic and molecular genetic analysis of multiple tumors obtained from members of the family with t(3;8) yielded similar results (Li et al., 1993). T h e question raised by these findings is whether the constitutional translocation has directly been involved in the inactivation of one allele of the tumor suppressor gene. Cytogenetically, distinct chromosomal sites are affected by translocations in the two families. One of the breakpoints was localized to chromosome 31314.2 (Wang and Perkins, 1984) and the other one to 3p13; thus, at least a subband of 3p14.1 separates the two breakpoints. Deletion mapping of the chromosome 3p region in tumors obtained from both families confirmed this finding. The loss of heterozygosity at D3S3 (PMS1-37), D3S42 (YNZ86.l), and D3S687 (CI3-528) loci was found in all tumors obtained from an individual with t(3;6), whereas tumors from individuals with t(3;8) retained the heterozygosity for these loci (unpublished observations, 1992). Therefore, it is unlikely that these translocations disrupt the same tumor suppressor gene. T h e loss of the translocated 3p segments in tumor cells may well be
MOLECULAR CYTOGENETICS OF RENAL CELL TUMORS
97
the result of instability of derivative chromosomes. Mitotic activity in normal kidney (probably during embryonal development) may simply result in a large number of cells with loss of the translocated 3p segment carrying one allele of the putative tumor suppressor gene. A mutational inactivation of the remaining allele in some of these cells may be instrumental in the development of multiple tumors. However, the functional effect of these chromosomal rearrangements could be clarified only when genes in proximity to the breakpoints become available for analysis. Recently, DNA probes were mapped around the chromosome 3p14.2 band in an effort to clone this breakpoint (van der Hout et al., 1991b; Yamakawa et al., 1992). ASSOCIATED WITH C. RENALCELLCARCINOMA HIPPEL-LINDAU DISEASE
VON
Von Hippel-Lindau (VHL) disease is an autosomally inherited disorder with predisposition to the development of tumor-like lesions and tumors in multiple organs (Lamiell et al., 1989; Maher et al., 1990). T h e symptoms are generally manifested between the ages of 20 and 40 years. The major lesions are retineal, cerebellar, brain stem, and spinal cord haemangioblastoma, renal, pancreatic, and epididemal cyts and pheochromocytoma. Multiple bilateral cysts of the kidney have been diagnosed in more than 50% of the gene carriers (Lamiell et al., 1989). Multiple nonpapillary renal cell carcinomas were found in 25-30% of these patients. There is no sex predominance for the development of renal cell carcinoma in von Hippel-Lindau patients. The gene responsible for the VHL phenotype was localized to chromosome 3p25-26 by familial linkage analysis (Seizinger et al., 1988; Hosoe et al., 1990). Until now, 46 renal cell carcinomas obtained from von HippelLindau patients were karyotyped (Decker et al., 1988; Goodman et al., 1990; Jordan et al., 1989; King et al., 1987; Kovacs and Kung, 1991; Kovacs et al., 1991a). All tumors showed the loss of chromosome 3p segment, which was the only karyotype change in 20 of 46 tumors. The RFLP analysis of multiple tumors from von Hippel-Lindau patients showed that chromosome 3p alleles inherited from the nonaffected parents were lost in each tumor (Kovacs and Kung, 1991; Tory et al., 1989). These data suggest a relationship between the loss of the wildtype allele of the VHL gene and the development of multiple lesions. Judging from the RFLP and chromosome analyses, loss of a large chromosomal segment of 3p13-pter inherited from nonaffected parent is the specific genetic change in renal cell carcinomas of von Hippel-Lindau patients. Because it is lost with the deleted chromosome 3p segment, it
98
GYULA KOVACS
was suggested that the VHL gene is the RCC suppressor gene (Tory et al., 1989). A deletion at 3p21 or 3p23 would effectively eliminate one allele of the VHL gene. However, none of the hereditary or sporadic tumors showed such a distal deletion suggesting that loss of more proximal regions is necessary before a cancer arises. It is more likely that the homozygous expression of VHL gene is responsible for the development of renal cysts and the inactivation of both alleles of the RCC gene for initiation of renal cell carcinomas. The clinical observation that the vast majority of renal cysts persist during life and do not develop a tumor supports this hypothesis (Maher et al., 1990). T h e genetic changes affecting other chromosomes are also similar to those found in sporadic renal cell carcinomas. Trisomy of the chromosome 5q22-qter segment was recorded in 54% of the tumors, in many cases as a result of a nonhomologous mitotic recombination between chromosome 3p and 5q. Both parental alleles of the chromosome 5q segment were involved in the genetic changes in multiple tumors in patients with von Hippel-Lindau disease (Kovacs and Kung, 1991). Alteration of chromosomes 6q, 8p, and 14q was found at lower frequency than in sporadic cases (Kovacs et al., 1991a). However, 24 of the 46 renal cell carcinomas were smaller than 1 cm in diameter, i.e., in an early stage of development, whereas almost all sporadic tumors were larger than 3 cm in diameter at the time of analysis.
D. NONHOMOLOGOUS MITOTICRECOMBINATION T h e loss of one allele of the putative RCC gene may occur at least three ways: nondisjunctional loss of the entire chromosome 3, deletion of a large chromosome 3p segment, and a nonhomologous mitotic recombination between chromosome 3 and other chromosomes. T h e translocation between chromosome 3p and 5q results in all cases in loss of one homologous chromosome 3p and partial trisomy of chromosome 5q segment. T h e RFLP analysis of tumor cells showed that both parental chromosomes 5 are retained and that one of them is partially duplicated (Kovacs and Kung, 1991). This result could be explained by the model of nonhomologous mitotic recombination (Fig. 1). In the late S phase or G2 phase, when two chromatids are available, a translocation between two nomhomologous chromatids may occur. After adjacent segregation, when daughter cells each receive a rearranged chromatid as well as normal chromatids, they are trisomic for one and monosomic for the other chromatid segment in exchange (Fig. 111). After alternate segregation, when both rearranged chromatids move into one of the daughter cells, the cell division results in a normal cell and in a cell with a balanced
MOLECULAR CYTOGENETICS OF RENAL CELL TUMORS
I
99
I1
Ill1
A B1 C D2
normal cell
partial trisomy 3p A
8182
nonhomologous chromatid exchange A 8 2 C D2
balanced translocation I
B 2 C D 1 .A _ _______I
partial monosomy 3p
partial monosomy 3p FIG. 1. Model of nonhomologous mitotic recombination in renal cell carcinomas (A and B represent chromosomes 3, whereas C and D are the partner chromosomes in exchange). Adjacent or alternate segregations after exchange between two nonhomologous chromatids (B2 and D2) lead to either unbalanced (I) or balanced (11) translocation. I, For recombination between chromosomes 3 and 5 and adjacent segregation, one daughter cells is trisornic for Sp and monosomic for 5q, whereas the other cell is monosomic for 3p and trisomic for 5q. 11, For involvement of chromosomes 3 and 8 and alternate segregation, o n e cell receives normal chromatids, whereas the other two rearranged chromatids result in a balanced translocation 3;8. This cell requires a nondisjunctional loss of derivative chromosome D2, to eliminate the 3p segment. The chromosome patter within dash-lined boxes fits the karyotype changes found in nonpapillary renal cell carcinomas (see Fig. 2) (Kovacs and Kung, 1991).
translocation (Fig. 1/11). Using this model, we can explain unbalanced translocations leading to loss of one and duplication of other chromosomal segments as well as balanced translocations. A partial monosomy of chromosome 3p and partial trisomy of chromosome 5q segments (Fig. 2A) or partial monosomy of both chromosome 3p and 8p segments
100
GYULA KOVACS
A 4
3
5
3
8
FIG. 2. Partial G-banded karyotypes showing the normal (left) and rearranged (right) chromosome 3 and chromosomes 5 and 8 involved in mitotic recombination. A, Unbalanced translocation (3;5) with two normal chromosomes 5. T h e breakpoint on chromosome 3 is at p13 (arrow). B, Translocation (3;8) and one copy of the normal chromosome 8. T h e breakpoint o n chromosome 3 p is marked by the arrow.
(Fig. 2B) are frequent cytogenetic findings in nonpapillary renal cell carcinomas. Of interest, karyotype changes corresponding to the other daughter with partial trisomy 3p and monosomy 5q was not yet detected in tumor cells. Likely, cells with duplication of chromosome 3p segment (three copies of the wild-type RCC gene) have less chance for a malignant transformation than normal cells or they are simply not viable. In families with a predisposition to renal cell carcinomas, a balanced translocation (3;6)and (3;s) is transmitted through the germline, and all cells of the kidney have this genetic alteration identical to the few precursor cells in sporadic cases having a balanced translocation after alternate segregation (Fig. 1/11).Therefore, a nondisjunctional loss of derivative chromosome 8 or 6 results in a high number of precursor cells in familial cases, and multiple nonpapillary renal cell carcinomas will develop. This model could be used for explanation of genetic changes occurring in cysts and tumors of von Hippel-Lindau patients. The VHL gene is mapped to the distal chromosomal 3p25-26 bands, whereas the RCC gene is located more proximally, likely somewhere around the 3p13-2 1 chromosomal region. A mitotic recombination, between chromosomes 3p13 and 5q22 in a patient carrying the VHL gene, results in cells with karyotype aberrations. One of the daughter cells has only the mutated allele of the VHL gene and one copy of the wild-type RCC gene (Fig. 111). This cell expresses the VHL phenotype, i.e., develops a renal cyst, which
MOLECULAR CYTOGENETICS OF RENAL CELL TUMORS
101
increases the number of cells with one copy of the wild-type RCC gene. A chromatid exchange between chromosomes 3 and 8 results in a daughter cell with balanced translocation 3;8 having two copies of the wildtype RCC gene, one copy of the wild-type and one of the mutated VHL gene. Nondisjunctional loss of the derivative chromosome 8 removing the wild-type VHL gene (and RCC gene) results in a cyst. A subsequent mutation in cells of cysts may lead to the development of nonpapillary renal cell carcinoma. Both types of segregation may occur at the same frequency by chance. T h e mechanism of alternate segregation requires more steps to eliminate one allele of the gene(s); therefore, it is less effective. The karyotypes of 18 tumors obtained from one patient with VHL disease support this theory. Seven tumors showed a karyotype alteration corresponding to the adjacent segregation (Fig. l/I), whereas loss of chromosome 3p suggesting an alternate segregation (Fig. 1/11) occurred in only two tumors (Kovacs et al., 1991a). T h e nonhomologous mitotic chromatid exchange is not limited to the alterations of the aforesaid chromosomes. This model could be used for the explanation of any other unbalanced translocations occurring in sporadic and hereditary renal cell carcinomas and in other types of tumors as well. A somatic crossing over is well documented in mammalian cells, and its cytological evidence, referred to as quadriradial chromosome configuration, is found in mitotic cells of cultured normal human tissues (German, 1964). Such quadriradials were interpreted as a consequence of interchange between two chromosomes with a point of exchange at apparently homologous sites. A somatic recombination via homologous chromatid interchange resulting in a partial uniparental disomy of chromosome 3 was demonstrated in a subclone of lymphoblastoid cells from a patient with Bloom's syndrome, a cancer-predisposing condition with genetic instability (Groden et al., 1990). Mitotic cells of individuals with Bloom's syndrome do accumulate an abnormally large number of mutations, many of which involve large chromosomal segments. All cytogenetic changes that characterize cells from individuals with chromosome instability syndromes are to be found, with much less frequency, in cells from normal individuals. A recombination between homologous and nonhomologous chromosomes is a frequent genetic event playing an important role in the evolution, and it occurs during individual development at a much higher frequency than thought (Steinmetz et al., 1987). Short tandemly repeated DNA sequences, which are present at about 1000 loci in the human genome, are implicated to be the sites of recombination (Wahls et al., 1990). A "genetic accident" at recombination sites might be a major
102
GYULA KOVACS
source of chromosomal translocations resulting in mosaicism in normal tissues (Tycko and Sklar, 1990).
E. SOMATIC MOSAICISMI N NORMAL KIDNEYTISSUE It was proposed that tumor cells arise from normal diploid cells and acquire karyotype alteration due to genetic instability (Nowell, 1986). Evidence for the role of genetic instability in tumor development comes from studies of chromosomal instability syndromes (Schroder, 1982). However, the vast majority of cancer patients have no inherited susceptibility to chromosomal alterations. The instability in these cases is the result of postzygotic events such as chromosomal mutation with random numerial or structural chromosomal changes in a subpopulation of somatic cells (Holliday, 1989). There is increasing evidence that chromosomal aberrations showing tissue-specific distribution may occur in normal human tissues (for a review see Hall, 1988). For example, only 1 of 85 lymphocytic cells of a renal cell cancer patient with constitutional translocation (3;6) showed an abnormal chromosomal set, whereas 15 of 62 kidney cells had structural and numerical karyotype alterations (Kovacs et al., 1989a). Nonrandom clonal chromosomal abnormalities, especially trisomy of chromosomes 7 and 10 as well as loss of the Y chromosome, occurs also in short-term cultures of normal kidney, lung, and brain tissues of patients with cancer (Elfving et al., 1990); Heim et al., 1989; Lee et al., 1987; Kovacs and Brusa, 1989). Structural or numerical karyotype alterations may occur up to 18% of metaphase cells of normal kidneys (Emanuel et al., 1992). Of interest, 2 of 2413 normal parenchymal renal cells obtained from patients with renal cell cancer have a deletion of the chromosome 3p segment (unpublished observations, 1991). These data suggest that a large number of cells may contain gross chromosomal alterations in phenotypically normal tissues. Taking into account that karyotyping is not sensitive enough to detect all genetic changes, the frequency of somatic mosaicism in normal tissues should be higher than generally assumed.
F. MODELOF DEVELOPMENT AND PROGRESSION OF NONPAPILLARY RENAL CELLCARCINOMA One allele of a putative tumor suppressor gene at chromosome 3p is inactivated by gross chromosomal alterations in 96% of nonpapillary renal cell carcinomas. As predicted by the Knudson model, the other
MOLECULAR CYTOGENETICS OF RENAL CELL TUMORS
103
allele should be inactivated by mutation before a tumor develops. Chromosome aberration arises during mitosis, and the cell division per se increases the risk of genetic errors of various kinds (Ames and Gold, 1990). There is a rapid, almost exponential cell proliferation during the embryonal and fetal period of life. When the kidneys have attained their normal size, the cells are replaced only to balance cell loss. Thus, the chance for gross chromosomal alterations is much higher during embryonal development than during cellular turnover and regeneration. Gene mutations occur independently from the cell cyclus, and the rate of gene mutations is time dependent. It is likely, that most chromosomal aberrations leading to somatic mosaicism arise during embryonal development, whereas most gene mutations are acquired during the postnatal period of life. Therefore, the following model for the development of sporadic and hereditary nonpapillary renal cell carcinomas is proposed. Genetic accidents during embryonal development result in mosaicism for various gross chromosomal aberrations, such as deletion 3p, nondisjunctional loss of chromosome 3, or mitotic recombination between 3p and 5q. A mutation at the remaining allele of the suppressor gene occurs in many cells during life. T h e homozygous inactivation of the RCC gene remains phenotypically silent in nonproliferating, differentiated tubular cells. When these cells are involved in cellular turnover or regeneration, they cannot stop to divide due to lack of function of suppressor gene. In addition, most of these cells have three copies of chromosome 5q22 segment or loss of the chromosome 6q, 8p, or other segments, which become an important genetic alteration when cells begin to proliferate. The genetic noise affecting the chromosome 3p segment in kidney cells of individuals with constitutional translocations is much higher than in sporadic cases. Therefore, all individuals carrying the germline translocation will develop multiple and/or bilateral renal cell carcinomas, if they live long enough (Li et al., 1993). Renal cell carcinoma in von Hippel-Lindau disease arises from preexisting cysts (Solomon and Schwarz, 1988). This developmental sequence is suggested by some clinical and histopathological data. Renal cysts are frequently detected in the second decade of the life, but renal cell carcinomas has not been reported in patients aged less than 20 years (Maher et al., 1990). T h e correlation between the number of cysts and tumors in kidneys from VHL gene carriers also support this hypothesis (unpublished observations, 1991). Both the VHL and RCC genes are located near the chromosomal region, which is deleted in nonpapillary renal cell carcinomas. The loss of this chromosomal segment during the embryonal development results in renal cyst. The growth of multiple
104
GYULA KOVACS
cysts increases the number of cells having only one homolog of the wildtype RCC gene, the mutational inactivation of which may result in the development of multiple renal cell carcinomas. According to the proposed model, a gross chromosomal deletion removing one allele of the RCC gene is the first genetic event in the vast majority of hereditary and sporadic nonpapillary renal cell carcinomas. We d o not know the number of cells affected by gross chromosomal alterations and/or mutations. Taking into account not only the genetic but also all epigenetic factors, an “optimal condition” for the development of one sporadic renal cell carcinoma is given in only 7 of 100,000 individuals. Most tumor cells are genetically unstable and acquire additional aberrations during clonal expansion. The accumulation of multiple genetic alterations and their association with the tumor phenotype from hyperplastic growth to frankly malignant tumor is well documented in colorectal carcinoma (Fearon and Vogelstein, 1990). T h e development and progression of nonpapillary renal cell carcinoma does not relate to an adenoma-carcinoma sequence. None of the known tumor suppressor genes (RB, APC, DCC, WT, p53) are affected. However, as discussed previously, alterations of at least six chromosomes are involved in the genetic changes of nonpapillary renal cell carcinomas. To determine the sequence of genetic alterations during clonal progression, we have divided renal cell carcinomas into groups corresponding to the number of karyotype alterations (Table 11). The frequency of specific aberrations was then determined for each group. This survey shows that loss of the chromosome 3p segment is the first genetic change. Trisomy of chromosome 5q22-qter region is the second genetic alteration, which arises in many tumors together with the loss of chromosome 3p. The third step is the monosomy of chromosome 14 in the majority of cases, followed by a monosomy of chromosomes 8 p and 9. T h e order of genetic alterations is TABLE I1 GENETIC CHANCES I N NONPAPILLARY RENALCELLCARCINOMAS Number of alterations
Karyotype changes (%)
-3p
+5q ~
1 2
3
4-5 >6
~
77
-
100 100 94 100
50 48 58
70
-6q ~
-8p
-9
-14q
8
-
Number of tumors
~~~~~~
-
8 13 17 24
15 4 8 41 42
6 50
8 52
53 75
13 24 23 17 24
105
MOLECULAR CYTOGENETICS OF RENAL CELL TUMORS
not constant, in some exceptional cases even the monosomy 8 or 9 might be the first visible karyotype change. A trisomy of chromosome 5q or monosomy of chromosomes 6q and 14q was not found as the only karyotype change. T h e loss of chromosome 3p and trisomy of 5q seqments occurs at the same frequency in carcinomas with and without metastatic growth (Table 111). This finding indicates that alterations of genes at chromosome 3p and 5q are associated with the development of renal cell carcinomas. A monosomy of chromosome 14q occurs in only 23% of tumors without metastatic growth, whereas it is found in 73% of tumors with metastasis. Recently, a similar association between loss of chromosome 14 and poor outcome in a subset of neuroblastoma was reported (Fong et al., 1992). It is likely that a gene at chromosome 14q is responsible for the alteration of tumor phenotype leading to a more aggressive, metastatic growth. In summary, nonpapillary renal cell carcinomas are characterized by genetic events affecting multiple chromosomal regions such as chromosome 3p, 5q, 6q, 8p, 9, and 14q. These data pinpoint a network of genes, the altered function of which is involved in the development and progression of these tumors. IV. Genetics of Papillary Renal Cell Tumors Papillary renal cell tumors account for approximately 10% of renal cancer. Histologically, they consist of papillary or tubulopapillary growth of small cuboidal cells with scanty cytoplasm or large columnar cells with eosinophilic or basophilic granular cytoplasm. There is a 6: 1 to 8: 1 preponderance of males over females. Papillary renal cell tumors have genetic changes that are distinct from those of nonpapillary carcinomas (Table I). Not only the chromosomes involved in karyotype alterations but also the genetic events are different. Even though nonpapillary renal cell carcinomas are marked by sequential losses of specific chromosomal regions, papillary renal cell tumors display trisomies such as trisomy of TABLE 111 GENETIC CHANGES ASSOCIATED WITH TUMOR ACCRESSIVENESS Karyotype changes (7%)
Metastatic growth
-3p
+5q
-6q
-8p
-9
no Yes
96 98
52 50
10 23
18 30
34
7
-14q
Number oftumors
30 73
74 26
106
GYULA KOVACS
chromosomes 3q, 7, 8, 12, 16, 17, and 20, with the exception of the loss of the Y chromosome.
A. GENETIC ALTERATIONS ASSOCIATED WITH PAPILLARY RENALCELLTUMORS 1 . Loss of the Y Chromosome
T h e loss of the Y chromosome is one of the specific karyotype changes occurring in more than 80% of tumors that arise in male patients. T h e papillary renal cell tumor develops preferentially in males, the ma1e:female ratio is 6: 1 to 8: 1. Nonpapillary renal cell carcinoma, which develops in the same organ at the same age of onset, has a missing Y chromosome in only 27% of the cases, and the ma1e:female ratio is 1.5: 1. T h e loss of the Y chromosome, in combination with trisomy of chromosomes 7 and 17, is the first visible karyotype change in papillary renal cell tumors (Kovacs et al., 1991b). The loss of Y chromosomespecific DNA sequences was confirmed by the RFLP analysis of tumor tissues (unpublished observations, 1992). How does the loss of the Y chromosome foster the development of papillary renal cell tumors? Gonadoblastomas occur exclusively in individuals with dysgenetic gonads having a Y chromosome (Verp and Simpson, 1987). It was postulated that a regulatory gene referred to as a gonadoblastoma locus at the Y chromosome (GBY gene) may act as an oncogene (Page, 1987). It is unlikely that alteration of the same gene is associated with the development of papillary renal cell tumors, because they are characterized by the loss of the Y chromosome. It is more likely that a tumor suppressor gene is localized at one of the homologous region on the X and Y chromosomes, a mutational and deletional inactivation of which is instrumental in the initiation of tumors. Peltomaki et al. (1991) suggested that the pseudoautosomal region is the site of specific genetic alterations in solid tumors. We are not able to confirm this finding in renal cell tumors (unpublished observations, 1992). Two of the papillary renal cell carcinomas that developed in females showed translocations involving the Xq22 band. The Xq22 and Ypl 1 chromosomal regions harbor homologous sequences and might be the loci of the tumor suppressor gene affected in papillary renal cell tumors. 2 . Polysomy of Chromosome 7
Trisomy or tetrasomy of chromosome 7 occurs in all papillary renal cell adenomas and in 75% of papillary carcinomas. Trisomy of chromosome 7 is one of the most common karyotype alterations occurring in
MOLECULAR CYTOGENETICS OF RENAL CELL TUMORS
107
various tumors and normal tissues. Trisomy 7 occurs in 18% in nonpapillary renal cell tumors. We do not know the genes that are affected by this gross chromosomal change. Polysomy of chromosome 7 and alteration of the EGFR gene has been implicated in the genetics of malignant glial tumors (Libermann et al., 1985). Overexpression of the EGFR gene was found in 60-93% of “renal cell carcinomas generally” (Freeman et al., 1989; Ishikawa et d.,1990; Weidner et al., 1990). Taking into account that trisomy of chromosome 7 occurs in approximately 30% of kidney cancers (all types together), a correlation between trisomy 7 and overexpression of the EGFR gene in papillary renal cell tumors could be excluded. Likely, the dosage alteration of another gene is the important event. Both hepatocyte growth factorlscatter factor (HGFISF) and its receptor (c-MET proto-oncogene) are mapped to chromosome 7. The HGF/SF has a mitogenic and morphogenic effect on renal tubular cells by activation of the c-MET protooncogene (Montesano et al., 1991; Nagaike et al., 1991). One can suggest that an autocrine growth stimulation involving the scatter factor and its receptor trigger the growth of cells having three copies of both genes. However, no data are available on the function of these genes in renal cell tumors.
3 . TriSomy of Chromosome 17 Trisomy of chromosome 17 occurs in all papillary renal cell adenomas and in 80% of the papillary renal cell carcinomas. The specific combination of trisomy of chromosomes 7 and 17 in the vast majority of tumors, with additional trisomy of chromosome 12 in some cases, suggests that the ERBB gene family, mapped to these chromosomes, might be involved in the genetic changes. T h e low expression of the ERBB-2IHER-2 gene was found in SO-SO% of the “renal cell carcinomas” (Freeman et al., 1989; Weidner et al., 1990). However, as with the EGFR gene and trisomy of chromosome 7, a correlation between the altered expression of ERBB-2 and trisomy of chromosome 17 could be excluded. The p53 tumor suppressor gene at the short arm of chromosome 17 is implicated in the genetics of nearly all types of tumors (de Fromentel and Soussi, 1992). One allele of the p53 gene is duplicated in all papillary renal cell tumors with trisomy 17 and the expression of p53 gene is 3-6 times higher in tumor tissues than in corresponding normal kidneys (unpublished observations, 1991).Because the overexpression of the p53 gene is characteristically (but not exclusively) associated with the presence of mutations in the coding sequences of the gene, we have analyzed 40 papillary renal cell tumors for mutation. However, PCR-SSCP analysis of the exons 2 to 11 of the p53 gene failed to detect any mutation. The RFLP analysis revealed the duplication of the same allele in multiple
108
GYULA KOVACS
papillary renal cell tumors from the same kidney. In another case, both tumors from the left kidney showed the duplication of one allele, whereas the other allele was duplicated in both tumors from the right kidney (unpublished observations, 1991). These data suggest that a somatic mutation o r imprinting and allelic dosage of genes (chromosomes) may be the initial alteration at chromosome 17 in papillary renal cell tumors. 4 . Trisomies Associated with Malignant Growth
Trisomy of chromosome 16 occurs in 62% of papillary renal cell carcinomas. T h e long arm of chromosome 16 harbors two genes, the uvomorulin and CAR genes, alteration of which is implicated in the aggressive growth of tumors. One allele of the uvomorulin (E-cadherin) gene is duplicated in carcinomas having trisomy 16. All but one papillary renal cell carcinomas showed a reduced expression of the uvomorulin gene, whereas in one tumor it was overexpressed. The lack of expression of the uvomorulin gene was shown to be associated with cell dissociation and metastatic tumor growth (Behrens et a!., 1989). Recently, a putative cell adhesion regulator ( C A R )gene was cloned and mapped to a chromosomal site near the uvomorulin gene locus (Pullman and Bodmer, 1992). Of interest, the loss of heterozygosity at this chromosomal region is associated with the malignant progression of various types of tumors (Carter et al., 1990; Sat0 et al., 1991; Tsuda et al., 1990). The frequent loss of constitutional heterozygosity at the same chromosomal region was detected in a subset of Wilms’ tumors as well (Maw et al., 1992). Trisomy of chromosome 12 occurs in 34% of papillary renal cell carcinomas. There is a frequent karyotype change (27%) in Wilms’ tumors as well. Polysomy of the short arm, i.e., i( 12p), and deletion of the long arm of chromosome 12 is a highly specific genetic alteration in male germ cell tumors (Atkin and Baker, 1983; Murty et al., 1992). A trisomy 12 occurs in over 80%of ovarian granulosa-stromal cell tumors, in most cases as the sole karyotype change (Fletcher et al., 1991). Trisomy 20 was found in 28% of papillary renal cell carcinomas. This chromosome aberration is relatively frequent in embryonal tumors such as rhabdomyosarcoma, hepatoblastoma, and Wilms’ tumor. Trisomy 8 occurs in 18% of papillary renal cell carcinomas, whereas the nonpapillary renal cell carcinomas show the loss of chromosome 8 sequences in 22% of the cases. Cytogenetic analysis of 32 papillary renal cell carcinomas revealed a trisomy of chromosome 3 in 6 tumors. Another 5 tumors showed a partial trisomy of the chromosome 3qll-qter segment due to an unbalanced translocation between chromosome 3 q l l and other chromosomes. Thus, a partial trisomy of chromosome 3ql l-qter sequences
MOLECULAR CYTOGENETICS OF RENAL CELL TUMORS
109
occurs in 34% of papillary renal cell carcinomas. This karyotype pattern could also be explained by the genetic mechanism of nonhomologous mitotic recombination. However, the recombinational breakpoint in papillary renal cell tumors is localized not at the short arm of chromosome 3, but at chromosome 3ql1, and the chromosomes in exchange are different. A duplication of the chromosome 3q sequences seems to be the important alteration, because only the daughter cell with trisomy of chromosome 3q and monosomy of the partner chromosome is seen in papillary renal cell carcinomas.
5 . Possible Functional Effect of Trisornies The identification of nonrandom trisomies in karyotypes of papillary renal cell tumors seems to provide evidence contrary to the known mechanism of mutational-deletional two-hits model of carcinogenesis. Trisomies, especially the combination of trisomies, are highly specific genetic changes in papillary renal cell tumors: none of the tumors showed a monosomy of such chromosomes. Trisomy of specific chromosomes is a common genetic change in experimental tumors and transformed rodent cells as well (Aldaz et al., 1989; Cowell, 1980; Cram et al., 1983). In such cases, preferential duplication of chromosomes carrying the mutant gene was shown (Bianchi et al., 1990; Bremmer and Balmain, 1990; Wirschubsky et al., 1984). There may be no contradiction between the two genetic mechanisms, i.e., the loss of chromosomal segments with the wild-type gene o r duplication of chromosomes carrying the mutant gene. The effect of mutation on cell proliferation-differentiation may be dependent on gene dosage (Klein, 1981). In cells that overcome the transacting restrictive control of the wild-type allele, the ratio might be one to zero when the wild-type allele is lost, or two to one or three to one when the mutant allele is duplicated. Another possible explanation is that modifier genes (imprinting genes), which are sensitive to gene dosage, are located on chromosomes 7 and 17, and other chromosomes involved in genetic changes of papillary renal cell tumors (Sapienza, 1990). B. ADENOMA-CARCINOMA SEQUENCE On the basis of karyotype alterations and histological-clinical characteristics, we are able to distinguish two groups of papillary renal cell tumors. Papillary renal cell adenomus are characterized by a combination of tri- o r tetrasomy of chromosome 7 and trisomy of chromosome 17 in each case (Table I). Until now, the cytogenetics of only 10 papillary renal cell adenomas have been published (for a review see Kovacs et al., 1991b).
110
GYULA KOVACS
Nine of the tumors were diagnosed in males, and the loss of the Y chromosome was detected in 7 of them. The size of benign papillary tumors varied between 2 mm and 5.5 cm in diameter. The cytogenetic finding that a constant combination of alteration of three chromosomes (-Y,+7,+ 17) occurs in small as well as in large tumors suggests a relative stability of these genetic changes during growth. Thus, the size of papillary renal cell tumors does not correlate with their biological behavior. Papillary renal cell carcinomas show genetic changes in addition to those of renal cell adenomas (Table I). Until now, karyotypes of 32 papillary renal cell carcinomas (27 tumors from males) were published (Carroll et d.,1987; DalCin et d.,1989; deJong et d.,1988; Kovacs, 1989; Kovacs et al., 1991b; Miles et al., 1988; Wolman et al., 1988). The size of the tumors varied between 7 mm and 19 cm in diameter. The most frequent alteration associated with the malignancy is trisomy of chromosome 16 followed by trisomy 12, partial trisomy 3q12-qter, trisomy 20, as well as trisomy 8 and monosomy 14. The association of these chromosomal alterations with malignant behavior suggests that these genetic changes are a prerequisite for an aggressive growth of papillary renal cell tumors. Renal cell adenomas may reach a large size without any sign of malignancy, but a malignant transformation accompanied by complex genetic changes may occur in small adenomas. It is not the size, but the accumulation of genetic alterations, that is associated with malignant behavior. C. PARENCHYMAL LESIONSASSOCIATED WITH PAPILLARY RENALCELLTUMORS For many years, multiple small papillary adenomas were described in kidneys of patients with renal cell carcinomas (Apitz, 1944; Cristol et al., 1946). Recently, a detailed histological analysis revealed 42 microscopic parenchymal lesions on average in kidneys with papillary renal cell carcinoma (Kovacs and Kovacs, 1993) and less than one (0.4) such alteration per kidney was detected in cases with nonpapillary renal cell carcinoma. T h e vast majority of these tubulo-papillary structures were intermingled with normal parenchymal elements. No strict border between the lesions and normal tissues or compression of surrounding normal parenchymal was seen; therefore, they do not fulfill the morphological criteria necessary for the diagnosis of an adenoma. These lesions resemble maturing and adenomatous nephrogenic rests, which are known to be associated with the development of Wilms’ tumor (Beckwith et al., 1990). These findings suggest that Wilms’ tumor in children and papillary renal cell
MOLECULAR CYTOGENETICS OF RENAL CELL TUMORS
111
tumors in adults develop from similar precursor lesions of ,embryonal origin. D. SHARED GENETIC CHANCES I N WILMS’ TUMOR AND PAPILLARY RENALCELLTUMOR Wilms’ tumor shares some genetic features with papillary renal cell carcinoma. Hyperdiploidy with nonrandom trisomies is the most common cytogenetic alteration in Wilms’ tumor (Kaneko et al., 1991). Hyperdiploidy with combination of specific trisomies is a characteristic of papillary renal cell tumors as well. Trisomies of chromosomes 3, 7,8, 12, 16, 17, and 20 are, in a variable degree, common in both types of tumor. T h e most intriguing data from our point of view are the karyotypic findings in three Wilms’ tumors published by Kaneko et al., (1991). Each tumor showed a combination of trisomy 7 and 17 as well as additional karyotype changes such as trisomy 12 and 20 among others. This combination of karyotype changes is pathognomonic for papillary renal cell carcinomas of adults. Whether these tumors represent an early transformation of nephrogenic rest into less differentiated papillary renal cell carcinomas in children or they are typical Wilms’ tumor is not yet known. The occurrence of persisting renal blastema and “Wilms’ tumor” in adults (Babaian et al., 1980; Scharfenberg and Beckman, 1984) as well as papillary adenoma in children (Stambolis, 1977) suggests that no strict age- and phenotype-defined border exists between childhood- and adult-type lesions. A genetic analysis of such exceptional cases is necessary to have more insight into the nature of these lesions. E. FROMDEVELOPMENTAL DISTURBANCES TO PAPILLARY RENALCELLCARCINOMAS Papillary renal cell tumor is an extraordinarily useful model for exploring the multistep nature of tumorigenesis from embryonal developmental disturbances to frankly malignant tumors of elderly patients. Since some basic knowledge regarding the normal nephrogenesis will be necessary for the following discussion, a quick sketch of some steps of the early stages of normal kidney development is provided here. The kidneys develop from mesodermal blastema. Their morphogenesis is triggered by inductive interactions between the epithelial “ureteric” bud and the mass of “nephrogenic mesenchyme.” As a result of bilateral inductive events, the epithelial ureteric bud will produce the collecting duct system, and the nephrogenic mesenchyme will go to develop a
112
GYULA KOVACS
normal nephron. T h e blastemal cells undergo an aggregation, vesiculation, and segmentation to form glomerular and tubular structures, and the latter makes an anastomosis with the “inducer”-collecting tubule. The renal architecture of such elements will be built up layer by layer from the inducable nephrogenic blastema, which remain detectable at the periphery of the renal lobes until the 36th week of pregnancy. Considering the complexity of this inductive communication system, it is no wonder that even in normal conditions sometimes errors occur. Thus, incompletely differentiated cells, referred to as nephrogenic rests, are left over from the fetal developmental phase. This is a fairly common event and may be subsequent to the alteration of gene(s), which control the proliferation and differentiation of blastemal cells. The Wilms’ tumor is thought to originate in cells of nephrogenic rests. Histological observations on multiple nephrogenic restlike parenchymal lesions in kidneys with papillary renal cell carcinomas suggest that the initial developmental sequence is shared by Wilms’ tumor of children and most of the papillary renal cell tumors of adults (Fig. 3) (Kovacs and Kovacs, 1993). T h e development of Wilms’ tumor is associated with the alteration of genes at chromosome 1lp13 and 1lp15 chromosomal region (for a review see Haber and Housman, 1992). The Wilms’ tumor gene (WTI)was
f
11P
/
-Y,+7,+17
+3q,+8.+12,+16,+20
carcinoma
FIG. 3. Model of development of papillary renal cell tumors. How nephrogenic rests develop and regress is not yet known. The predicted developmental sequence for Wilms’ tumor is the mutational inactivation of genes at chromosome 1 Ip in nephrogenic rests (Haber and Housman, 1992). It is proposed that many of the nephrogenic rests persist during life, and some with genetic alteration of a putative tumor suppressor gene at X-Y chromosomes and with combined trisomy of chromosome 7 and 17 develop benign papillary adenoma. A papillary renal cell carcinoma develops (in most cases) in cells having additional genetic changes such as trisomy of chromosomes 3q, 8, 12, 16, and 20 (G. Kovacs et al., 1992).
MOLECULAR CYTOGENETICS OF RENAL CELL TUMORS
113
recently cloned from the chromosome l l p 1 3 region (Call et al., 1990; Gessler et al., 1990). The involvement of WTI in the urogenital organ development and its mutation in Wilms’ tumors was shown (Haber and Housman, 1992). We d o not know whether these genetic changes are associated with the maturation block of cells in nephrogenic rests. When the mutation of genes at chromosome 1l p region is responsible for the persistence of blastemal cells in kidneys, and both Wilms’ tumor and papillary renal cell tumor are developed from the precursor lesions, the chromosome 1l p region should be altered in papillary renal cell tumors as well. The lack of alteration at chromosome l l p region in papillary renal cell tumors, however, excludes the possibility that gene(s) in this region are involved in the development of the precursor nephrogenic rests (G. Kovacs et al., 1992). The lack of linkage between familial Wilms’ tumor development and chromosome 1 1p 13 or 11p 15 region support this hypothesis (Grundy et al., 1988; Huff et al., 1988). Recently, the loss of heterozygosity for markers at chromosome 16q13-22 was also implicated in the etiology of some Wilms’ tumors (Maw et al., 1992). However, a genetic linkage of familial Wilms’ tumor predisposition to this chromosomal region has been ruled out (Huff et al., 1992). It is more likely that a yet unidentified gene(s) is responsible for the maturation arrest and “overproduction” of embryonal blastemal cells, the inherited or somatic mutation of which may result in a faulty induction stimuli of differentiation. T h e continuous cell growth, after cessation of normal kidney development, results in nephrogenic rests. Presumably, most of the nephrogenic rests regress in early ages (Beckwith et al., 1990). When cells of the nephrogenic rests acquire alteration of genes at chromosome l l p (Haber and Housman, 1992) and/or other chromosomal region such as 16q, 6, 12 (Kaneko et al., 1991; Maw et al., 1992), a Wilms’ tumor develops. When cells of nephrogenic rests acquire genetic changes such as alteration of a putative tumor suppressor gene at the homologous region of the X and Y chromosomes and trisomy of chromosomes 7 and 17, they proliferate slowly and undergo some degree of differentiation. After many decades of slow proliferation, they may result in papillary renal cell adenoma and, subsequently to additional genetic changes, in papillary renal cell carcinoma. The early stage of tumorigenesis, until nephrogenic rests develop, is thought to be shared by Wilms’ tumor and papillary renal cell tumor. However, the alterations at chromosome 1l p in Wilms’ tumor and at the Y chromosome in papillary renal cell tumors indicate clearly that they have distinct genetics. The alteration of the Y chromosome is one of the initial genetic changes in papillary renal cell tumors, but it is a rare karyotype alteration in Wilms’ tumors. There is a strong 6: 1 to 8: 1 male
114
GYULA KOVACS
preponderance for the development of papillary renal cell tumors, whereas Wilms’ tumor develops equally in both sexes. Some families with inherited predisposition to the development of Wilms’ tumor were described, but an inherited form of papillary renal cell tumor has not yet been reported. As previously suggested, the inherited genetic alteration in families with Wilms’ tumor yields a large number of maturation arrested cells, i.e., nephrogenic rests, which highly increase the risk for Wilms’ tumor development. Most of these children will be cured. However, not only Wilms’ tumors but also the nephrogenic rests will be “cured” by the therapy; therefore, no more precursor lesions will be left over for late papillary renal cell tumor development. F. PAPILLARY RENALCELLTUMORS WITH TRANSLOCATION INVOLVING THE Xp11.2 BREAKPOINT Recently, a new cytogenetic subtype of papillary renal cell carcinoma having a translocation (X; l)(pl1.2;q21) was described by Meloni et al. (1993). Of interest, one of the four cases had a trisomy of both chromosomes 7 and 17 and another one a trisomy of chromosome 17, which are the characteristic genetic changes for the main group of papillary renal cell tumors. Two additional renal cell tumors of adults with translocation between the X chromosome and chromosome 1 have been described. Both tumors were characterized by identical translocation involving the Xpll.2 and 1p34.3 chromosomal bands (Yoshida et al., 1986; Kovacs et al., 1987b). Renal cell carcinoma is rare in children, comprising less than 7% of renal cell tumors in individuals under 21 years of age (Raney et al., 1983). Only two cases have been analyzed cytogenetically. DeJong et al. (1986) karyotyped a renal cell cancer obtained from a 2.4-year-old boy. Each tumor cell has a 46,Y,t(X: l)(p11.2;q21) karyotype. The second case was reported by Tomlinson et al. (1991). A 17-month-old boy had multiple lymph node metastasis at the time of surgery. The tumor cells displayed a 46,Y,t(X;17)(pl1.2;q25)karyotype, while the karyotype of peripheral blood lymphocytes was normal. In addition to the common breakpoint, tumors belonging to this genetic group have two common features. First, each tumor had a trabecular/papillary growth pattern of extremely large clear cells, of which the cellular phenotype is rarely seen among papillary renal cell tumors with trisomy of chromosome 17. Second, all cases were observed in male patients. T h e common breakpoint at chromosome Xpll.2 in tumors suggests that this region may contain a tumor suppressor gene, the
MOLECULAR CYTOCENETICS OF RENAL CELL TUMORS
115
alteration of which is associated with the development of this subtype of papillary renal cell tumors. The chromosomal band Xp 11.2 is implicated also in the development of synovial sarcoma, another mesodermally derived epithelial malignancy that is characterized by a specific translocation (X;18)(pl1.2;q11.2) (Turc-Carel et al., 1987), and in the genetic changes of uterine leiomyomas as well (Mark et al., 1990). V. Renal Oncocytoma Renal oncocytoma is a benign tumor of the kidney. The wellcircumscribed tumor is composed of acinar-arranged, large eosinophilic cells. Electron microscopic studies showed that cells of oncocytomas are densely packed with mitochondria. Only few data on the cytogenetics of renal oncocytomas are published so far. Some of these cases represents other types of renal tumors such as "oncocytic" or chromophobe renal cell carcinomas and show loss of chromosome 3 or other karyotype changes characteristic for such tumors (Psihramis et al., 1986; Dobin et al., 1992). Employing RFLP techniques, no loss of heterozygosity at chromosome 3p sequences was found in renal oncocytomas (Brauch et al., 1990; G . Kovacs, unpublished observations, 1993). Karyotypes of only a few renal oncocytomas are available for the present survey. From these data, however, we can outline some characteristics of the genetic alterations.
A. CHROMOSOMAL ALTERATIONS A subset of renal oncocytomas displays a mixed population of cells with normal and abnormal karyotypes showing clonal chromosomal abnormalities (Crotty et al., 1992; Dobin et al., 1992; Kovacs et al., 1987c, 1989b). Seven renal oncocytomas have balanced or unbalanced translocations involving different chromosomes. Translocations (7;9)(q22;q13) and (X,13)(qll;p13) were found in one case (Kovacs et al., 1987c), a translocation (1; 13)(q21;q34) was found in another case, and a translocation (20;?)(p13;?),in the third tumor (Kovacs et al., 1989b). Dobin et al. (1992) described a renal oncocytoma showing a translocation (14; 17)(pll;p13). Three oncocytomas shared a common breakpoint at chromosome 1lq13. One of them had a 46,XY,t(9;1l)(p23;q13)karyotype (Walter et al., 1989). Presti et al. (1991) described an oncocytoma with a 46,XX,t(5;1 l)(q35;q13) karyotype. T h e third case (Kerman et al., 1991) displayed a 45,XY,-1,-1 l,+der(ll)t(l;ll)(pll;qll) chromosomal pattern. Reevaluation of the published karyotype of the latter case suggests the breakpoint to be at chromosome 1lq13. Thus, the chromosome
116
GYULA KOVACS
1lq13 is a common breakpoint in these cases. This chromosomal band represents a “hot spot“ region containing genes, the rearrangement or amplification of which is associated with the development and progression of various type of tumors (for a review see Lammie and Peters, 1991). A paracentric inversion of chromosome 11 in benign parathyroid adenomas places the PTH gene from 1lp15 adjacent to the PRAD I gene at llq13 leading to its elevated expression. The amplification of the chromosome 1lq13 region containing the INT2, H S T I , PRAD 1, and EMS1 genes has been found in breast cancer, transitional cell carcinoma of the urinary bladder, non-small cell lung carcinoma, and squamous cell carcinoma of the head and neck region. It is unknown whether these genes are involved in the initiation and progression of renal oncocytomas. T h e chromosome 1 lq13 band harbors two fragile sites, namely FRAllA, a rare folk acid type, and FRAllH, a common amphidicolin type fragile site, which might also be instrumental in the translocations. However, the alteration of a gene(s) at this chromosomal region is more likely than the consistent involvement of one of the many fragile sites. Recently, Crotty et al. (1992) have suggested that coincident loss of the Y chromosome and chromosome 1 as the sole abnormality marks a subset of renal oncocytomas. Until now, seven oncocytomas having the 44,X,-Y,-1 karyotype have been reported (Crotty et al., 1992; Jordan et al., 1992; Meloni et al., 1992; Miles et al., 1992; Psihramis et al., 1988). Crotty et al. (1992) also described a renal oncocytoma with a balanced translocation 15;21 and loss of the Y chromosome and another one with monosomy 22 as the sole karyotype change. Thus, renal oncocytomas comprise a cytogenetically heterogeneous group of tumors. An end-to-end fusion or telomere-to-centromere rearrangement of chromosomes was observed in some of the renal oncocytomas. Telomeric association of chromosomes occurs in normal cells from patients with ataxia teleangiactasia (Hayashi and Schmid, 1975), in senescent human fibroblasts (Benn, 1976) and also in tumors (Kovacs et al., 1987a). The role of these genetic changes in the tumor development as well as the correlation between the heterogeneous chromosomal DNA and constant mitochondria1 DNA alterations, if any, is not yet established. B. MITOCHONDRIAL DNA ALTERATIONS T h e genetic mechanism underlying the accumulation of mitochondria in renal oncocytomas is not known. The high number of mitochondria may reflect some disturbances in regulation of mitochondrial division.
MOLECULAR CYTOGENETICS OF RENAL CELL TUMORS
117
The D-loop region is the site where protein-DNA interaction occur directing the mitochondrial DNA replication (Zeviani et al., 1989). All proteins involved in the replication are thought to be encoded by nuclear genes. Molecular analysis of mitochondrial DNA from oncocytoma tissues revealed an altered restriction pattern after Hinff digestion (Kovacs et al., 1989b; Welter et al., 1989), which might reflect a mutation in one of the restrictions fragment (Singh et al., 1987). Whether mitochondrial DNA alteration is responsible for the continuous replication resulting in the enormously high number of mitochondria in tumor cells or the chromosomal DNA alterations at 1lq13, 1, and Y, remains to be analyzed.
VI. Chromophobe Renal Cell Carcinoma Recently, a new phenotypical variant of renal cell tumor, referred to as chromophobe renal cell carcinoma (RCC), was described (Thoenes et al., 1988). Chromophobe renal cell carcinoma is characterized by cells having a pale, fine reticular cytoplasm. By electron microscopic analysis, chromophobe cells display pathognomic cytoplasmic vesicles and a variable number of mitochondria with an altered morphology. A. CHROMOSOMAL ABERRATIONS Until now, the chromosome analysis of three chromophobe RCCs has been published (Kovacs et al., 1988b; Kovacs and Kovacs, 1992). These tumors showed an unusually low chromosome number between 34 and 39. In two fully karyotyped cases, the common loss of chromosomes 1,2, 6, 10, 13, 17, and 21 has been found. It was suggested that these karyotype changes reflect a clonal selection of cells with specific chromosomal losses. Telomeric association between different chromosomes as well as pulverization of multiple chromosomes were also found. Both types of alterations may lead to extreme loss of chromosomes in descendent cells. The RFLP analysis using chromosome 3p-, 5q-, 17p-, and 17q-specific DNA probes showed allelic changes that have not been seen in other types of renal cell tumor (A. Kovacs et al., 1992). To establish the genetic changes of chromophobe renal cell carcinomas, we have employed the comparative genomic hybridization (CGH) technique. Hybridization of DNA extracted from 17 chromophobe renal cell carcinomas to normal metaphasis chromosomes revealed a constant loss of chromosomes 1, 2, 6, 10, 13, 17, and 21 (Table I). The loss of chromosome 1 occurred in 100% of the cases, whereas the loss of other chromosomes occurred in
118
GYULA KOVACS
76-95% of tumors. This finding is very unusual, and we do not have any interpretation for such an extreme loss of chromosomes in a constant combination.
B. MITOCHONDRIAL DNA ALTERATIONS Chromophobe renal cell carcinomas are 'characterized by a variable number of normal and morphologically altered mitochondria as well as by cytoplasmic vesicles, which are thought to originate from mitochondria (Bonsib and Lager, 1990; G. Kovacs, unpublished observations, 1992). T h e mitochondrial DNA may be extensively rearranged in some of the chromophobe renal cell carcinomas (A. Kovacs et al., 1992). Whether these changes affect the function of mitochondrial genes, and if so, what genes are involved, is not yet known. Whether the extensive alteration of chromosomal DNA in chromophobe renal cell carcinomas results in the alteration of mitochondrial DNA and in the abnormal morphology of mitochrondria also remains to be established. Further molecular biological studies will be necessary to understand the complexity of genetic events affecting the chromosomal and mitochondrial DNA in chromophobe renal cell carcinomas and renal oncocytomas as well.
VII. Conclusions T h e identification of specific chromosomal and mitochondrial DNA alterations affecting sites of yet unknown genes in renal cell tumors offers a fascinating series of questions to scientists interested in diverse areas. There is a high rate of cell proliferation and complex process of cell differentiation and maturation during the fetal period of life. Mature cells of the nephron reach the stage of terminal differentiation at the 36th weeks of gestation, when the proliferation genes become suppressed. When the kidneys have attained their normal size, the cells are replaced only to balance loss. Papillary renal cell tumors arise from cells corresponding to the developmental stage before the 36th week of gestation, whereas nonpapillary renal cell carcinomas develop from differentiated tubular cells. Likely, genetic alterations at chromosome 3q, '7, 8, 12, 16, and 17 and at the X and Y chromosomes in papillary tumors pinpoint a network of genes the normal function of which is important in the embryonal and early fetal stage of kidney development. The complex genetic changes at chromosome 3p, 5q, 6q, 8p, 9, and 14q in nonpapillary renal cell carcinomas may identify loci of genes, which control the limited cell proliferation such as normal cellular turnover or
MOLECULAR CYTOGENETICS OF RENAL CELL TUMORS
119
regeneration. Cloning these genes and establishing their function in the morphogenesis and regeneration of nephrons may help researchers to understand how distinct renal cell tumors develop and progress. This will be a formidable task, since there are many genes involved in the regulation of growth, morphogenesis, and differentiation of the kidney. Future studies will establish, especially through the definition of gene alterations, the true nature of distinct types of renal cell tumors and, as a result, will determine the prognosis because it is through the use of such biological information that new approaches to treatment will be developed. As soon as this happens, the determination of the genetic status of renal cell tumors will be an essential component of the oncologicalpathological service. ACKNOWLEDGMENTS This study was supported by grants from the German Research Council; the Department of Health and Human Services under Contract N01-CO-74102 with Program Resources, Inc.; the EMBO; and the Japanase Foundation for Promotion of Cancer Research. I am grateful to Drs. B. Beckwith, W.F. Bodmer, G. Klein, A. Knudson, and Y. Nakamura for stimulating discussions.
REFERENCES Aldaz, C. M., Trono, D., Larcher, F., Slaga, T. J., and Conti, C. J. (1989). Mol. Carcinog. 2, 22-26. Ames, B. N., and Gold, L. S. (1990). Sczence 249, 970-971. Anglard, P., Tory, K., Brauch, H., Weiss, G. H., Latif, F., Merino, M. J., Lerman, M. I., Zbar, B., and Linehan, W. M. (1991). Cancer Res. 51, 1071-1077. Apitz, K. (1944). Virchows Arch. 311, 328-359. Atkin, N. B., and Baker, M. C. (1983). Cancer Genet. Cytogenet. 10, 199-204. Babaian, R. J., Skinner, D. G., and Waisman, J. (1980). Cancer45, 1713-1719. Beckwith, J. B., Kiviat, N. B., and Bonadio, J. F. (1990). Pedzatr. Pathol. 10, 1-36. Behrens, J., Mareel, M. M., Van Roy, F. M., and Birchmeyer, W. (1989).J. Cell Biol. 108, 2435-2447. Benn, P. A. (1976). Am. J . Hum. Genet. 28,465-473. Bennington, J. L., and Labscher, F. A. (1968). Cancer 21, 1069-1071. Bergenheim, U., Nordenskjold, M., and Collins, V. P. (1989). Cancer Res. 49, 1390-1396. Berger, C. S., Sandberg, A. A,, Todd, I. A. D., Pennington, R. D., Haddad, F. S., Hecht, B. K., and Hecht, F. (1986). Cancer Genet. Cytogenet. 23, 1-24. Bianchi, A. B., Aldaz, C. M., and Conti, C. J. (1990). Proc. Natl. Acad. Scz. U.S.A.87,6902-
6906.
Bonsib, S. M., and Lager, D. L. (1990). Am. J. Surg. Pathol. 14, 260-267. Brauch, H., Tory, K., Linehan, W. M., Weaver, D. J., Lowell, M. A,, and Zbar, B. (199O).J. Vrol. 143, 622-624. Bremmer, R., and Balmain, A. (1990). Cell 61, 407-417. Call, K. M., Glaser, T., Ito, C. Y., Buckler, A. J., Pelletier, J., Haber, D. A., Rose, E. A,, Kral, A., Yeger, H., Lewis, W. H., Jones, C., and Housman, D. E. (1990). Cell 60,509-520.
120
GYULA KOVACS
Carroll, P. R., Murty, V. V. S., Reuter, V., Jhanwar, S., Fair, W. R., Whitmore, W. F., and Chaganti, R. S. K. (1987). Cancer Genet. Cytogenet. 26, 253-259. Carter, B. S., Ewing, C. M., Ward, W. S., Treiger, B. F., Aalders, T. W., Schalken, J. A,, Epstein, J. I., and Isaacs, W. B. (1990). Proc. Natl. Acad. Scz. U.S.A. 87, 87518755. Cohen, A. J., Li, F. P., Berg, S., Marchetto, D. J., Tsai, S., Jacobs, S. C., and Brown, R. S. (1979). N . Engl. J. Med. 301, 592-595. Cowell, J. K. (1980).JNCI, J. Natl. Cancer Inst. 65, 955-959. Cram, L. S., Bartholdi, M. F., Ray, F. A., Travis, G. L., and Kraemer, P. M. (1983). Cancer Res. 43,4828-4837. Cristol, D. S., McDonald, F. R., and Immell, F. L. (1946).J. Urol. 55, 18-27. Crotty, T. B., Lawrence, K. M., Moertel, C. A., Bartelt, D. H., Batts, K. P., Dewald, G. M., and Jenkins, R. B. (1992). Cancer Genet. Cytogenet. 61,61-66. DalCin, P., Li, F. P., Prout, G. R., Huben, R. P., Limon, J., Ferti-Passantonopoulou, A,, Richie, J. P., and Sandberg, A. A. (1988). Cancer Genet. Cytogenet. 35, 41-46. DalCin, P., Gaeta, J., Huben, R., Li, F.P., Prout, G. R., and Sandberg, A. A. (1989).A m . J . Clin. Pathol. 92, 408-414. Decker, H.-J. H., Neumann, H. P. H., Walter, T. A., and Sandberg, A. A. (1988). Cancer Genet. Cytogenet. 33, 59-65. d e Fromentel, C. C., and Soussi, T. (1992). Genes, Chromosomes Cancer 4, 1-15. deJong, B., Molenaar, I. M., Leeuw, J, A., Idenberg, V. J. S., and Osterhuis, J. W. (1986). Cancer Genet. Cytogenet. 21, 165-169. deJong, B., Oosterhuis, J. W., Idenburg, V. J. S., Castedo, S. M. M. J., Dam, A., and Mensink, H. J. A. (1988). Cancer Genet. Cytogenet. 30, 53-61. Dobin, S. M., Harris, C. P., Reynolds, J. A., Coffield, K. S., Klugo, R. C., Peterson, R. F., and Speights, V. 0. (1992). G a e s , Chromosomes Cancer 4, 25-31. Elfving, P., Lundgren, R., Cigudosa, J. C., Limon F., Mandahl, N., Kristofferson, U., Heim, S., and Mitelman, F. (1990). Cytogenet. Cell Genet. 53, 123-125. Emanuel, A., Szucs, S., Weier, H. U. G., and Kovacs, G. (1992).Genes, ChromosomesCancer 4, 75-77. Erlandsson, R., Bergenheim, U. S. R., Boldog, F., Marcsek, Z., Kunimi, K., Lin, B. Y.-T., Ingvarsson, S., Castresana, J. S., Lee, W.-H., Lee, E., Klein, G., and Sumegi, J. (1990). Oncogene 5 , 1207- 1211. Erlandsson, R., Boldog, F., Persson, B., Zabarovsky, E. R., Allikmets, R. L., Sumegi, J., Klein, G., and Jornvall, H. (1991). Oncogene 6 , 1293-1295. Fearon, E. R., and Vogelstein, B. (1990). Cell 61, 759-767. Fletcher, J. A., Gibas, Z., Donovan, K., Perez-Atayde, A., Genest, D., Morton, C. C., and Lage, J. M. (1991).J. Pathol. 138, 515-520. Fong, C., White, P. S., Peterson, K., Sapienza, C., Cavenee, W. K., Kern, S. E., Vogelstein, B., Cantor, A. B., Look, A. T., and Brodeur, G. M. (1992). CancerRes. 52, 1780-1785. Freeman, M. R., Washecka, R., and Chung, L. W. K. (1989). Cancer Res. 49,6221-6225. German, J. (1964).Science 144, 298-301. Gessler, M., Poustka, A., Cavenee, W., Neve, R. L., Orkin, S. H., and Bruns, G. A. P. (1990). Nature (London) 343, 774-778. Goodman, M. D., Goodman, B. K., Lubin, M. B., Braunstein, G., Rotter, J. I., and Schreck, R. R. (1990). Cancer 65, 1150-1 154. Groden, J., Nakamura, Y., and German, J. (1990).Proc. Natl. Acad. Sn. U.S.A. 87, 43154319. Groden, J., Thliveris, A., Samowitz, W., Carlson, M., Gelbert, L., Albertsen, H., Joslyn, G., Stevens, J., Spirio, L., Robertson, M., Sargeant, L., Krapcho, K., Wolff, E., Burt, R.,
MOLECULAR CYTOCENETICS OF RENAL CELL TUMORS
121
Hughes, J. P., Warrington, J., McPherson, J., Wasmuth, J., Le Paslier, D., Abderahim, H., Cohen, D., Leppert, M., and White, R. ( 1 9 9 1 ) . Cell 66, 589-600. Grundy, P., Koufos, A., Morgan, K., Li, F. P., Meadows, A., and Cavenee, W. K. (1988). ’ Nature (London) 336, 374-376. Haber, D. A., and Housman, D. E. (1992). Adv. Cancer Res. 59, 41-68. Hall, J. (1988). Am. J. Hum. Genet. 43, 355-363. Hayashi, K., and Schmid, W. (1975). Humungenetih 30, 135-141. Heini, S., Mandahl, N., Jin, Y., Strombald, S., Lindstrom, S., Salford, L. G., and Mitelman, F. (1989). Cytogenet. Cell Genet. 52, 136- 138. Holliday, R. (1989). Trendc G m t . 5, 42-45. Hosoe, S., Brauch, H., Latif, F., Glenn, G., Daniel, L., Bale, S., Choyke, P., G r i n , M., Oldfield, E., Berman, A., Goodman, J., Orcutt, M. L., Hampsch, K., Delisio, J., Modi, W., McBridge, W., Anglard, P., Weiss, G., Walther, M. M., Linehan. W. M., Lerman, M. I., and Zbar, B. (1990). Genomic.7 8, 634-640. Huff, V., Compton, D., Chao, L., Strong, L., Geiser, C., and Saunders, G. (1988). Nature (London) 336, 377-378. Huff, V., Reeve, A. E., Leppert, M., Strong, L. C., Douglass, E. C., Geiser, C. F., Li, F. P., Meadows, A,, Callen, D. F., Lenoir, G., and Saunders, G. F. (1992). Cancer Res. 52, 6117-6120. Ishikawa, J., Maeda, S., Umezu, K., Sugiyama, T., and Kamidono, S. (1990). In2.J. Cancer 45, 1018-1021. Javadpour, N. (1984). I n “Cancer of the Kidney” (N. Javadpour, ed.), pp. 5-13. ThiemeStratton, New York. Jordon, D. K., Patil, S. R.,Divelbiss, J. E., Vemuganti, S., Headley, C., Waziri, M. H., and Gurll, N. J. (1989). Cancer Genet. Cytogenet. 42, 227-241. Kaneko, Y., Homa, C., Meki, N., Sakurai, M., and Hata, J. (1991). Cancer Res. 51, 59375942. Kerman, S. L., Heritage D. W., and Hefter, L. G. (1991). Cancer Genet. Cytogenet. 56, 134135. King, C. R., Schimke, R. N., Arthur, T., Davoren, B., and Collins, D. (1987). Cancer Genet. Cytogenet. 27, 345-348. Kinzler, K. W., Nilbert, M. C., Su, L.-K., Vogelstein, B., Bryan, T. M., Levy, D. B., Smith, K. J., Preisinger, A. C., Hedge, P., McKechnie, D., Finniear, R., Markham, A,, Grofen, J., Boguski, M. S., Altschul, S. F., Horii, A,, Ando, H., Miyoshi, Y., Miki, Y., Nishisho, I., and Nakamura, Y. (1991a). Science 253, 661-665. Kinzler, K. W., Nilbert, M. C., Vogelstein, B., Bryan, T. M., Levy, D. B., Smith, K. J., Preisinger, A. C., Hamilton, S. R., Hedge, P., Markham, A., Carlson, M., Joslyn, G., Groden, J., White, R., Miki, Y., Miyoshi, Y., Nishisho, I., and Nakamura, Y. (1991b). Science 251, 1366-1370. Klein, G. ( 198 1). Nature (London) 294, 3 13-3 18. Knudson, A. G. (1987). Adu. Viral. Oncol. 7 , 1-17. Kovacs, G. (1989).A m . J . Pathol. 134, 27-34. Kovacs, G. (199O).J. Cancer Res. Clzn. Oncol. 116, 318-323. Kovacs, G. (1 993). Histopathology 22, 1-8. Kovacs, G., and Brusa, P. (1988). Hum. Genet. 60, 99-101. Kovacs, G., and Brusa, P. (1989). Cancer Genet. Cytogenet. 37, 289-290. Kovacs, G., and Frisch, S. (1989). Concer Res. 49, 651-659. Kovacs, A., and Kovacs, G. (1992). Genes, Chromosomes Cancer 4, 267-268. Kovacs, G., and Kovacs, A. (1993).J. Urol. Pathol. (in press). Kovacs, G., and Kung, H. (1991). Proc. Nutl. Acad. Sci. U.S.A. 88, 194-198.
122
GYULA KOVACS
Kovacs, G., Muller-Brechlin, R., and Szucs, S. (1987a). Cancer Genet. Cytogenet. 28,363-366. Kovacs, G . ,Szucs, S., DeRiese, W., and Baumgartel, H. (1987b). 1nt.J. Cancer 40, 171-178. Kovacs, G., Szucs, S., Eichner, W., Maschek, H. J. Wahnschaffe, V., and DeRiese, W. ( 1 9 8 7 ~ )Cancer . 59, 2071-2077. Kovacs, G., Erlandsson, R., Boldog, F., Ingvarsson, Muller-Brechlin, R., Klein, G., and Sumegi, J. (1988a). Proc. Natl. Acad. Sci. U.S.A. 85, 1591-1595. Kovacs, G., Soudah, B., and Hoene, E. (1988b). Cancer Genet. Cytogenet. 31, 211-215. Kovacs, G., Brusa, P., and DeRiese, W. (1989a). Int. J . Cancer 43, 422-427. Kovacs, G., Welter, C., Wilkens, L., Blin, N., and DeRiese, W. (1989b). Am. J. Pathol. 134, 967-97 1. Kovacs, G., Wilkens, L., Papp, T., and DeRiese, W. (1989~).JNCI, J. Natl. Cancer Inst. 81, 527-530. Kovacs, G., Emanuel, A., Neumann, H. P. H., and Kung, H. (1991a). Genes, Chromosomes Cancer 3, 256-262. Kovacs, G . , Fuzesi, L., Emanuel, A., and Kung, H. (1991b). Genes, Chromosomes Cancer 3, 249-255. Kovacs, A,, Storkel, S., Thoenes, W., and Kovacs, G. (1992). J. Pathol. 167, 273-277. Kovacs, G., Kiechle-Schwarz, M., Scherer, G., and Kung, H. (1992). Cell. Mol. B i d . 38,5962. LaForgia, S., Morse, B., Levy, J., Barnea, G., Cannizzaro, L.A., Li, F. P., Nowell, P.C., Boghosian-Sell, L., Click, J., Weston, A., Harris, C. C., Drabkin, H., Patterson, D., Croce, C. M., Schlessinger, J., and Huebner, K. (1991). Proc. Natl. Acad. Sci. U.S.A. 88, 5036-5040. Lamiell, J. M., Sabarar, F. G., and Hsia, Y. E. (1989). Medzctne (Baltimore) 68, 1-29. Lammie, G. A., and Peters, G. (1991). Cancer Cells 3, 413-420. Latif, F., Tory, K., Modi, W. S., Graziano, S. L., Gamble, G., Douglas, J., Heppel-Parton, A. C., Rabbitts, P. H., Zbar, B., and Lerman, M. I. (1992). Genes ChromosomesCancer5,119127. Lee, J. S., Pathak, S., Hopwood, V., Tomasevic, B., Mullins, T. D., Baker, F. L., Spitzer, G., and Neidhart, J. A. (1987). Cancer Res. 47, 6349-6352. Li, F. P., Marchetto, D. J., and Brown, R. S. (1982). Cancer Genet. Cytogenet. 7 , 271-273. Li, F. P., Decker, H.-J. H., Zbar, B., Stanton, V. P., Kovacs, G., Seizinger, B. R., Aburatani, H., Sandberg, A. A., Berg, S., Hosoe, S., and Brown, R. S. (1993).Ann. Intern. Med.118, 106-1 11. Libermann, T. A,, Nusbaum, H. R., Razon, N., Kris, R., Lax, I., Soreq, H., Whittle, N., Waterfield, M. D., Ullrich, A,, and Schlessinger, J. (1985). Nature (London) 313, 144147. Limon, J., Mrozek, K., Heim, S., Elfving, P., Nedoszytko, B., Babinska, M., Mandahl, N., Lundgren R., and Mitelman, F. (1990). Cancer Genet. Cytogenet. 49, 259-263. Maher, E. R., Yates, J. R. W., Harries, R., Benjamin, C., Harris, R., Moore, A. T., and Ferguson-Smith, M. A. (1990). Q.J. Med. 77, 1151-1163. Maloney, K. E., Norman, R. W., Lee, C. L. Y., Millard, 0. H., and Welch, J. P. (1991).J. Urol. 146,692-696. Mark, J., Havel, G., Grepp, C., Dahlenfors, R., and Wedell, B. (1990). Cancer Genet. Cytogenet. 44, 1-13. Maw, M., Grundy, P. E., Millow, E. J., Eccles, M. R., Dunn, R. S., Smith, P. J., Feinberg, A. P., Law, D. J., Paterson, M. C., Telzerow, P. E., Callen, D. F., Thomson, A. D., Richards, R. I., and Reeve, A. E. (1992). Cancer Res. 5 2 , 3094-3098. Meloni, A. M., Sandberg, A. A., and White, R. D. (1992). Cancer Genet. Cytogenet. 61, 108109.
MOLECULAR CYTOGENETICS OF RENAL CELL TUMORS
123
Meloni, A. M., Dobbs, R. M., Pontes, J. E., and Sandberg, A. A. (1993). Cancer Genet. Cytogenet. 65, 1-6. Miles, J., Michalski, K., Kouba, M.. and Weaver, D. J. (1988). Cancer Genet. Cytogenet. 34, 135-142. Miller, Y. E., Minna, J. D., and Gazdar, A. F. (1989).j. Clin. Invest. 83, 2120-2124. Montesano, R., Matsumoto, K., Nakamura, T., and Orci, L. (1991). Cell 67, 901-908. Morita, R., Saito, S., Ishikawa, J., Ogawa, O., Yoshida, O., Yamakawa, K., and Nakamura, Y. (1991). Cancer Res. 51, 5817-5820. Murty, V. V. V. S., Houldsworth, J., Baldwin, S., Reuter, V., Hunziker, W., Besmer, P., Bosl, G., and Chaganti, R. S. K. (1992). Proc. Natl. Acad. Sci. U.S.A. 89, 11006-1 1010. Nagaike, M., Hirao, S., Tajima, H., Noji, S., Taniguchi, S., Matsumoto, K., and Nakamura, T. (1991).J. Biol. Chem. 266, 22781-22764. Nowell, P. C. (1986). Cancer Res. 46, 2203-2207. Ogawa, O., Habuchi, T., Kakehi, Y., Koshiba, M., Sugiyama, T., and Yoshida, 0. (1992). Cancer Res. 52, 1881-1885. Page, D. C. (1987). Developmat, Suppl. 101, 151-155. Pathak, S., Strong, L. C., Ferrell, R. E., and Trindade, A. (1982). Science 217, 939940. Peltomaki, P., Lothe, R., Berresen, A,, Fossa, S. D., Brogger, A,, and d e la Chapelle, A. (1991). Int. J. Cancer 47, 518-522. Presti, J. C., Rao, P. H., Reuter, V. E., Li, F. P., Fair, W. R., and Jhanwar, S. C. (1991). Cancer Res. 51, 1544-1552. Psihramis, K. E., Althausen, A. F., Yoshida, M. A., Prout, G. R., and Sandberg, A. A. (1986).J. Urol. 136, 891-895. Psihramis, K. E., DalCin, P., Dretler, S. P., Prout, G. P., and Sandberg, A. A. (1988).J. Urol. 139,585-587. Pullman, W. E., and Bodmer, W. F. (1992). Nature (London) 356, 529-532. Rabbitts, P., Bergh, J., Collins, F., and Waters, J. (1990) Genes, Chromosomes Cancer 2, 231238. Raney, R. B., Palmer, N., Suzow, W. W., Baum, E., and Ayala, A. (1983). Med. Pediatr. Oncol. 11,91-98. Sapienza, C. (1990). Mol. Carcinog. 3, 118-121. Sato, T., Akiyama, F., Sakamoto, G., Kasumi, F., and Nakamura, Y. (1991). CancerRes. 51, 5794-5799. Scharfenberg, J. C., and Beckman, E. N. (1984). Hum. Pathol. 15, 791-793. Schroder, T. M. (1982). Cytogenet. Cell Genet. 33, 119-132. Seizinger, B. R., Rouleau, G. A., Ozelius, L. J., Lane, A. H., Farmer, G. E., Lamiell, J. M., Haines, J., Yuen, J. W. M., Collins, D., Majoor-Krakauer, D., Bonner, T., Matthew, C., Rubenstein, A,, Halperin, J., McConkie-Rosell, A,, Green, J. S., Trofatter, J. A ,, Ponder, B. A., Eierman, L., Bowmer, M. I., Schimke, R., Oostra, B., Aronin, N., Smith, D. I., Drabkin, H., Waziri, M. H., Hobbs, W. J., Martuza, R. L., Conneally, P. M., Hsia, Y. E., and Gusella, J. F. (1988). Nature (London) 332, 268-2653, Shimizu, M., Yokota, J., Mori, N., Shuin, T., Shinoda, M., Terada, M., and Oshimura, M. (1990). Oncogene 5, 185-194. Singh, G., Neckelman, N., and Wallace, D. C. (1987). Nature (London) 329, 270-272. Solomon, D., and Schwarz, A. (1988). Hum. Pathol. 19, 1072-1079. Stambolis, C. (1977). Virchows Ai-ch. .4 Pathol. Anat. Histol. 375, 267-272. Steinmetz, M., Uematsu, Y., and Fischer Lindahl, K. (1987). Trends Genet. 3, 7-10. Sutherland, G. R., and Hecht, F. (1985). “Fragile Sites on Human Chromosomes.”, Oxford University Press, New York.
124
GYULA KOVACS
Tajara, E. H., Berger, C. S., Hecht, B. K., Gemmill, R. M., Sandberg, A. A., and Hecht, F. ( 1988). Cancer Genet. Cytogenet. 31, 75-82. Teyssier, J. R..and Ferre, D. (1990). Cancer Genet. Cytogenet. 45, 197-205. Thoenes, W., Storkel, S., Rumpelt, H. J., Moll, R., Baum, H. P., and Werner, S. (1988). J. Pathd. 155,277-287. Tomlinson, G. E., Nisen, P. D., Timmons, C. F., and Schneider, N. (1991). Cancer Genet. Cytogenet. 57, 1 1- 17. Tory, K., Brauch, H., Linehan, M., Barba, D., Oldfield, E., Filling-Katz, M., Seizinger, B., Nakamura, Y., White, R., Marshall, F. F., Lerman, M. I., and Zbar, B. (1989). JNCIJ. Natl. Cancer Inrt. 81, 1097-1 101. Tsuda, H., Zhang, W., Shimosoto, Y., Yokota, J., Terada, M.,Sugimura, T., Miyamura, T., and Hirohashi, S. (1990). Proc. Natl. Acad. Sci. U.S.A. 87,6791-6794. Turc-Carel, C., DalCin, P., Limon, J.. Rao, U., Li, F. P., Corson, J. M., Zimmerman, R., Parry, D. M., Cowan, J. M., and Sandberg, A. A. (1987). Proc. Natl. Acad. Sci. U.S.A. 84, 1981- 1985. Tycko, B., and Sklar, J. (1 990). Cancer Cells 2, 1-8. van der Hout, A., van der Vlies, P., Wijmenga, C., Li, F. P., Oosterhuis, J. W., and Buys, C. H. C. M. (1991a). Genomics 11, 537-542. van d er Hout, A., Brown, R. S., Li, F. P., and Buys, C. H. C. M. (1991b). Cancer Genet. Cytogenet. 51, 121-124. Verp, M. S., and Simpson, J. L. (1987). Cancer Genet. Cytogenet. 25, 191-218. Vogelstein, B., Fearon, E. R., Kern, S. E., Hamilton, S. R.,Preisinger, A. C., Nakamura, Y., and White, R. (1989). Science 444, 207-2 1 1 . Wahls, W. P., Wallace, L. J., and Moore, P. D. (1990). Cell 60, 95-103. Walter, T. A., Berger, C. S. and Sandberg, A. A. (1989). Cancer G a e t . Cytogenet. 43, 15-34. Wang, N., and Perkins, K. L. (1984). Cancer Genet. Cytogenet. 11, 479-481. Warrington, J. A., Bailey, S. K., Armstrong, E., Aprelikova, O., Alitalo, K., Dolganov, G. M., Wilcox, A. S., Sikela, J. M., Wolfe, S. F., Lovett, M., and Wasmuth, J. J. (1992). Genomics 13, 803-808. Weidner, U., Peter, U., Strohmeyer, T., Hussnatter, R., Ackerman, R., and Sies, H. 1990). Cancer Res. 50,4504-4509. Welter, C., Kovacs, G., .Seitz, G., and Blin, N. (1989). Genes, Chromosomes Cancer 1, 79-82. Wirschubsky, D. J., Wiener, F., Spira, J., Siimegi, J., and Klein, G. (1984).Znt. J. Cancer 33, 477-48 1 . Wolman, S. R.,Camuto, P. M., Golimbu, M., and Schinella, R. (1988). CancerRes. 48,28902897. Yamakawa, K., Morita, R., Takahashi, E., Hori, T., Ishikawa J., and Nakamura, Y. (1991). Cancer Res. 51,4707-47 1 1 . Yamakawa, K., Takahashi, E., Murata, M., Okui, K., Yokoyama, S., and Nakamura, Y. (1992). Genomics 14,412-416. Yoshida, M. A., Ohyashiki, K., Ochi, H., Gibas, Z., Pontes, E., Prout, G. R., Huben, R., and Sandberg, A. A. (1986). Cancer Res. 46,2139-2147. Yunis, J. J., and Soreng, A. L. (1984). Science 226, 1199-1204. Zbar, B., Brauch, H., Talmadge, C., and Linehan, M. (1987). Nature (London) 327, 721724. Zeviani, M., Servidei, S., Gellera, C., Bertini, E., DiMauro, S., and DiDonato, S. (1989). Nature (London) 339, 309-3 1 I .